Search

Background: Deep learning (DL)-based organ segmentation is increasingly used in radiotherapy, yet voxel-wise DL uncertainty maps are rarely presented to clinicians. Purpose: This study assessed how DL-generated uncertainty maps impact radiation oncologists during manual correction of prostate radiotherapy DL segmentations. Methods: Two nnUNet models were trained by 10-fold cross-validation on 434 MRI-only prostate cancer cases to segment the prostate and rectum. Each model was evaluated on 35 independent cases. Voxel-wise uncertainty was calculated using the SoftMax standard deviation (n=10) and visualized as a color-coded map. Four oncologists performed segmentation in two steps: Step 1: Rated segmentation quality and confidence using Likert scales and edited DL segmentations without uncertainty maps. Step 2 ($\geq 4$ weeks later): Repeated step 1, but with uncertainty maps available. Segmentation time was recorded for both steps, and oncologists provided qualitative free-text feedback. Histogram analysis compared voxel edits across uncertainty levels. Results: DL segmentations showed high agreement with oncologist edits. Quality ratings varied: rectum segmentation ratings slightly decreased overall in step 2, while prostate ratings differed among oncologists. Confidence ratings also varied. Three oncologists reduced segmentation time with uncertainty maps, saving 1-2 minutes per case. Histogram analysis showed 50% fewer edits for step 2 in low-uncertainty areas. Conclusions: Presenting DL segmentation uncertainty information to radiation oncologists influences their decision-making, quality perception, and confidence in the DL segmentations. Low-uncertainty regions were edited less frequently, indicating increased trust in DL predictions. Uncertainty maps improve efficiency by reducing segmentation time and can be a valuable clinical tool, enhancing radiotherapy planning efficiency.

We consider the quantum Friedmann equations which include one-loop vacuum fluctuations due to gravitons and scalar field matter in a FLRW background with constant $\epsilon=-{\dot{H}}/{H^2}$. After several field redefinitions, to remove the mixing between the gravitational and matter degrees of freedom, we can construct the one loop correction to the Friedmann equations. Due to cosmological particle creation, the propagators needed in such a calculation are typically infrared divergent. In this paper we construct the graviton and matter propagators, making use of the recent construction of the infrared finite scalar propagators calculated on a compact spatial manifold in \cite{Janssen:2008px}. The resulting correction to the Friedman equations is suppressed with respect to the tree level contribution by a factor of $H^2/m_p^2$ and shows no secular growth., Comment: 37 pages, 1 figure. Revised version, including a proper treatment of the infrared sector

We consider the one-loop effective action due to gravitons in a FLRW background with constant epsilon=-(dH/dt)/H^2. By expanding around epsilon=0 (corresponding to an expansion around de Sitter space), we can study how the deviation from de Sitter space effects the quantum corrected Friedmann equations. We find that, at zeroth order in epsilon, one-loop effects induce only a finite shift in the coupling constants. At linear order in epsilon there is however a divergent contribution to the equations of motion. This contribution leads to a nontrivial term in the renormalized equations that depends logarithmically on H and thus cannot be absorbed in local counterterms. We find that deviations due to this term are unobservably small. Our study shows that quantum effects in quasi de Sitter space can be fundamentally different then in de Sitter space, albeit in the case under consideration the effect is unobservably small., Comment: 25 pages, 3 figures

We construct the scalar and graviton propagator in quasi de Sitter space up to first order in the slow roll parameter $\epsilon\equiv -\dot{H}/H^2$. After a rescaling, the propagators are similar to those in de Sitter space with an $\epsilon$ correction to the effective mass. The limit $\epsilon\to 0$ corresponds to the E(3) vacuum that breaks de Sitter symmetry, but does not break spatial isotropy and homogeneity. The new propagators allow for a self-consistent, dynamical study of quantum back-reaction effects during inflation., Comment: 23 pages

We consider quantum effects of a massive antisymmetric tensor field on the dynamics of de Sitter space-time. Our starting point is the most general, stable, linearized Lagrangian arising in nonsymmetric gravitational theories (NGTs), where part of the antisymmetric field mass is generated by the cosmological term. We construct a renormalization group (RG) improved effective action by integrating out one loop vacuum fluctuations of the antisymmetric tensor field and show that, in the limit when the RG scale goes to zero, the Hubble parameter -- and thus the effective cosmological constant -- relaxes rapidly to zero. We thus conclude that quantum loop effects in de Sitter space can dramatically change the infrared sector of the on-shell gravity, making the expansion rate insensitive to the original (bare) cosmological constant., Comment: 32 pages, 2 eps figures

We consider the linearized nonsymmetric theory of gravitation (NGT) within the background of an expanding universe and near a Schwarzschild mass. We show that the theory always develops instabilities unless the linearized nonsymmetric lagrangian reduces to a particular simple form. This form contains a gauge invariant kinetic term, a mass term for the antisymmetric metric-field and a coupling with the Ricci curvature scalar. This form cannot be obtained within NGT. Based on the linearized lagrangian we know to be stable, we consider the generation and evolution of quantum fluctuations of the antisymmetric gravitational field (B-field) from inflation up to the present day. We find that a B-field with a mass m ~ 0.03(H_I/10^(13)GeV)^4 eV is an excellent dark matter candidate., Comment: 9 pages, 1 figure. Based on two talks by the authors at the 2nd International Conference on Quantum Theories and Renormalization Group in Gravity and Cosmology (IRGAC) 2006, Barcelona

We consider the linearized nonsymmetric theory of gravitation (NGT) within the background of an expanding universe and near a Schwarzschild metric. We show that the theory always develops instabilities unless the linearized nonsymmetric lagrangian reduces to a particular simple form. This theory contains a gauge invariant kinetic term, a mass term for the antisymmetric metric-field and a coupling with the Ricci curvature scalar. This form cannot be obtained within NGT. Next we discuss NGT beyond linearized level and conjecture that the instabilities are not a relic of the linearization, but are a general feature of the full theory. Finally we show that one cannot add ad-hoc constraints to remove the instabilities as is possible with the instabilities found in NGT by Clayton., Comment: 29 pages